Proc. Natl Acad. Sci. USA Vol. 78, No. 2, pp. 766-769, February 1981 Biochemistry

Immunological identification of a common precursor to arginine and neurophysin II synthesized by in vitro translation of bovine hypothalamic mRNA (signal sequence/polyprotein/processing/core glycosylation/tunicamycin) HARTWIG SCHMALE AND DIETMAR RICHTER* Institut ffir Physiologische Chemie, Abteilung Zellbiochemie, Universitit Hamburg, Martinistrasse 52, 2 Hamburg 20, Federal Republic of Germany Communicated by Nathan 0. Kaplan, October 17, 1980 ABSTRACT mRNA from membrane-bound polysomes of bo- MATERIALS AND METHODS vine was translated in an mRNA-dependent cell- free system from reticulocyte lysate or wheat germ. The trans- Bovine (Np I) and neurophysin II (Np II) and lation products were identified by immunoprecipitation with an- rabbit antisera to Np I and Np II were purchased from Bio- tibodies to either neurophysin H or arginine vasopressin followed products (Brussels); according to the supplier the antisera by sodium dodecyl sulfate/polyacrylamide gel electrophoresis. An against Np II were passed over a Np I-Sepharose column to im- immunoreactive polypeptide was obtained with an apparent Mr prove the specificity of anti-Np II. The same was done for anti- of 21,000. Sequential immunoprecipitation studies indicated that the Mr 21,000 product is a common precursor to neurophysin II Np I. The supplier claimed that this procedure reduced cross- and arginine vasopressin. The specificity of the immunoprecipi- reactivity of the neurophysin antisera to less than 0.5% as tation was demonstrated by competition with excess amounts of judged by radioimmunoassay. Both these and the rabbit anti- unlabeled neurophysin H or arginine vasopressin; little or no com- serum against AVP (Ferring, Kiel) were precipitated with 28 petition was observed with unlabeled neurophysin I or . g of ammonium sulfate per 100 ml. According to Ferring, there Processing experiments with microsomal membranes from dog is no crossreaction between anti-AVP and oxytocin or lysine pancreas or tunicamycin-treated ascites tumor cells showed that the Mr 21,000 polypeptide is the prepro form. It was converted vasopressin and less than 30%o with ; anti-AVP does not to a pro form with Mr 19,000 suggesting a pre sequence of ap- discriminate between oxidized and reduced AVP. Hypothalamic proximately 15 amino acids. The Mr 19,000 polypeptide was core- mRNA was isolated as reported (13). The cell-free wheat germ glycosylated to an apparent Mr of 23,000, indicating that the neu- system was prepared according to published procedures (13). rophysin II-arginine vasopressin precursor is a glycopolypeptide. The reticulocyte lysate assay system came from New England Nuclear. EDTA-stripped microsomal membrane fractions from Oligopeptides such as the bacterial antibiotics gramicidin and dog pancreas (14) were kindly provided by B. Dobberstein tyrocidin are synthesized in a nonribosomal fashion on polyen- (Heidelberg). Microsomal membranes from ascites tumor cells zyme templates (1, 2). Reports that oligopeptides of hypothal- (15, 16) treated with tunicamycin (Lilly) and immobilized mem- amus from mammals also might be synthesized by a nonribo- branes from Staphylococcus aureus (17) were prepared as somal mechanism are still controversial (3-5). In vitro biosyn- reported. thesis of the hypothalamic nonapeptide vasopressin and oxytocin has not yet been resolved. Although Sachs (6) sug- gested more than a decade ago that vasopressin may be syn- RESULTS thesized by a ribosome-dependent mechanism via a precursor Identification of the Immunoreactive Products. mRNA iso- polypeptide, only recently a tentative glycopolypeptide pre- lated from detached polysomes of bovine hypothalami was cursor with an apparent M, of 20,000 has been isolated from the translated in a cell-free system from reticulocyte lysate or wheat hypothalamus of rats and mice (7-9). This precursor gave rise germ in the presence of [3S]cysteine. Purified antibodies to a vasopressin-like peptide after trypsinization (7). In addition raised against neurophysin II (anti-Np II) or arginine vasopres- the precursor crossreacted with antibodies raised against one sin (anti-AVP) were used to identify the translation products. of the hypothalamic cysteine-rich (8, 9) which is The immunoreactive products precipitated with S. aureus im- known to transport vasopressin to its storage site in the posterior munoabsorbant were analyzed by sodium dodecyl sulfate/ pituitary (10). polyacrylamide gel electrophoresis. With anti-Np II, two prod- If the precursor hypothesis is valid it should be possible to ucts with apparent Mr of 21,000 and 18,000 were obtained, demonstrate biosynthesis of a vasopressin-neurophysin pre- whereas with anti-AVP only the Mr 21,000 product was found cursor directly by translation of hypothalamic mRNA in cell- (Fig. 1, lanes 1 and 2). The Mr 18,000 product was barely de- free systems. Although a number of groups have reported the tectable when mRNA was translated in the cell-free wheat germ synthesis of neurophysin precursors in cell-free systems (11-13), system. At present it is not possible to decide whether the Mr identification of vasopressin-like sequences within the reported 18,000 product represents another discrete neurophysin II-like precursors has been lacking. The present communication re- polypeptide (18) or whether it is a premature Mr 21,000 poly- ports mRNA-directed synthesis of a common precursor to ar- peptide derived from a pre-termination event. ginine vasopressin (AVP) and neurophysin II and its processing The specificity of the immune reactions was demonstrated and core glycosylation; specific antibodies were used for identification. Abbreviations: Np I, bovine neurophysin I; anti-Np I, antibodies against Np I raised in rabbits; Np II, bovine neurophysin II; anti-Np II, anti- The publication costs of this article were defrayed in part by page charge bodies against Np II raised in rabbits; AVP, arginine vasopressin; anti- payment. This article must therefore be hereby marked "advertise- AVP, antibodies raised against AVP in rabbits. nent" in accordance with 18 U. S. C. §1734 solely to indicate this fact. * To whom reprint requests should be addressed. 766 Downloaded by guest on October 2, 2021 Biochemistry: Schmale and Richter Proc. Natl. Acad. Sci. USA 78 (1981) 767

Anti-Np II: + _- - + -

Anti-Np I:

Anti-AVP: + - ± + + ± + _-

Competing peptides cr > W" 0. > U) z C a z . a -

Mr. x 10 s x 103 c 21 21 18 18 __ .:... o - 16.5

1 2 3 4 5 6 7 8 9 10 11

FIG. 1. Gel electrophoresis of the Np ll-AVP precursor after synthesis and immunoprecipitation. Bovine hypothalamic mRNA (0.5 Mug) prepared from detached polysomes and purified by oligo(dT)-cellulose chromatography (13) was translated in the wheat germ system (lanes 3 and 4) or the reticulocyte lysate system (remaining lanes) in a total volume of 25 Ml [10 ,l of lysate, 2 ,1 of translation cocktail, 0.5 Ml of 32.5 mM Mg(OAc)2, 2 ,l of 1 M KOAc, and 20 MCi of [35S]cysteine]. After incubation at 370C the mixture was diluted with 3 vol of cold 10 mM Na phosphate, pH 7.6/ 1 mM EDTA/1% Triton X-100 containing 200 units of Antagosan (Behring, Marburg) per ml. For immunoprecipitation, 10 Mug of anti-Np I or anti- Np II IgG or 20 Mg of anti-AVP IgG (1.4 A280 units = 1 mg) were added and incubated at 40C for 20 hr. Nonspecific aggregates were removed by centrifugation at 15,000 x g for 15 min; the supernatant fraction was subjected to indirect immunoprecipitation with S. aureus immunoabsorbant (13, 17). After washing by three cycles of centifugation (8000 x g, 2 min) and resuspension in 20 mM Tris'HCl, pH 7.6/1 mM EDTA/500 mM NaCl/ 0.5% Triton X-100, the antigen-antibody complexes were dissociated from the staphylococci by boiling for 3 min in 62.5 mM Tris-HCl, pH 6.8/5% (wt/vol) sodium dodecyl sulfate/10% (wt/vol) glycerol/4% (vol/vol) 2-mercaptoethanol and subsequently subjected to sodium dodecyl sulfate/15% polyacrylamide gel electrophoresis. The dried slab gels were fluorographed. [For the total translation products (prior immunoprecipitation), see ref. 13.] In this and the following fluorogram, lanes were derived from different gels and aligned according to the position of marker (a, oval- bumin, Mr 43,000; b, carbonic anhydrase, 30,000; c, soy bean trypsin inhibitor, 20,000; d, a-lactalbumin, 14,400). Where indicated, the immuno- precipitation was carried out in the presence of excess (10 Mg per assay) of unlabeled competing peptide [bovine serum albumin (BSA), AVP, Np II, oxytocin (OT), or luteinizing -release hormone (LHRH)].

by competition experiments. Excess AVP effectively inhibited munoprecipitation step, implying that the Mr 21,000 polypep- binding of the Mr 21,000 product to AVP specific antibodies tide is a common precursor to Np II and AVP. In some exper- (Fig. 1, lane 5); no inhibition was observed with excess Np II Table 1. Sequential immunoprecipitation of the (lane 6), oxytocin or LHRH On the (lane 7), (lane 8). other hand Np II-AVP precursor excess Np II (lane 9) but not AVP (lane 10) inhibited binding of the Mr 21,000 and 18,000 polypeptides to anti-Np II. Np I Total cpm in Np ll-AVP was significantly less potent as a competitor than Np II although Antibodies added precursor at high concentrations Np I was also inhibitory (not shown). 1st step 2nd step 1st step 2nd step II did not Apparently, anti-Np discriminate completely be- Anti-Np II Anti-AVP 1192 192 tween two reverse the neurophysins. Similarly, in the experi- Anti-AVP 1044 290 ment anti-Np I crossreacted with the Mr 21,000 and 18,000 Anti-Np II precursors, although it predominantly reacted with a polypep- Translation in the reticulocyte lysate system, immunoprecipitation, tide of Mr 16,500 (lane 11). The Mr 16,500 polypeptide has been and gel electrophoresis were carried out as described in Fig. 1. For se- identified as the precursor to Np I and oxytocin (19). quential immunoprecipitation, 25 tld of the translation assay mixture was subjected to an initial immunoprecipitation step with anti-Np II. To determine whether both anti-Np II and anti-AVP were The immunoprecipitation was collected by centrifugation; the super- recognizing the same polypeptide, sequential immunoprecip- natant fraction of this step was supplemented with anti-AVP. A sim- itation was performed. The immunoprecipitate obtained with ilar experiment was performed in the reverse order using a 25-1.l Ali- anti-Np II was sedimented and the' supernatant fraction sub- quot of the translation mixture and starting the immunoprecipitation sequently was supplemented with anti-AVP. A similar experi- with anti-AVP; the immunoreactive material was precipitated by cen- ment was performed in the reverse order. The results (Table trifugation and the supernatant fraction supplemented with anti-Np II. The immunoreactive material was analyzed by sodium dodecyl sul- 1) show that the Mr 21,000 product $iad been removed from the fate gel electrophoresis and fluorography. The radioactive polypeptide assay mixture by the first immunoprecipitation step; little or corresponding to the Np II-AVP precursor was cut out of the gels and no immunoreactive material was detectable in the second im- measured in a liquid scintillation counter. Downloaded by guest on October 2, 2021 768 Biochemistry: Schmale and Richter Proc. Natl. Acad. Sci. USA 78 (1981)

No membranes With membranes

I

Anti-Np II: + _ + + _ - + - - + +

Anti-AVP: + - - + + - + + _ _ + + -

b -- Mr x 10-3 Mr 23 _ O a- a- X 10-3 c- 21 ~ 18 -_ I - 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14

FIG. 2. Synthesis and processing of the Np ll-AVP precursor in the absence and presence of microsomal membranes and sequential immu- noprecipitation. Translation of bovine hypothalamic mRNA in the reticulocyte lysate system, immunoprecipitation, and gel electrophoresis were carried out as described in Fig. 1. Lanes 3-12: 5 A260 units of dog pancreas microsomal membranes per ml of assay mixture. Lanes 13 and 14: 3 A260 units of microsomal membranes from tunicamycin-treated ascites tumorcells (15) were added per ml of assay mixture prior to incubation. Lanes 4 and 6: immunoprecipitation in the presence of excess Np II or AVP, respectively (10 Ag). Lanes 7-10 sequential immunoprecipitation: 1st step, anti-Np II (lane 7); 2nd step, anti-AVP (lane 8); 1st step, anti-AVP (lane 9); 2nd step, anti-Np II (lane 10). Lanes 11 and 12: Np II-AVP precursor synthesized in the presence of dog pancreas microsomal membranes, immunoprecipitated with anti-Np II (lane 11) or anti-AVP (lane 12), bound to concanavalin A-Sepharose columns and eluted with 0.2 M a-methyl mannoside (24). The Mr 19,000 and 18,000 products were not retained on the columns and appeared in the sugar-free eluate (data not shown). The Mr 23,000 product was bound to the column quantitatively and more than 80% was recovered after elution with a-methyl mannoside. For Mr markers, see Fig. 1.

iments, in particular when the first immunoprecipitation was The Mr 23,000 product could be identified as the glycosylat- with anti-AVP, the second step yielded minor immunoreactive ed pro form in two ways. First, the Mr 23,000 polypeptide, but material with anti-Np II (see also Fig. 2, lanes 9 and 10). Ap- neither the 19,000 nor the 21,000 polypeptide, was retained on parently, this result was due to an excess of the Mr 21,000 prod- a concanavalin A-Sepharose affinity column. The core-glyco- uct over anti-AVP and could be overcome by using higher con- sylated Mr 23,000 product could be eluted from the column with centrations of anti-AVP. 0.2 M a-methyl mannoside (Fig. 2, lanes 11 and 12). Second, Processing and Core Glycosylation of the Precursor. The when translation of hypothalamic mRNA was performed in the existence of a common precursor to neurophysin II and to the presence of a microsomal membrane fraction derived from tun- oligopeptide hormone AVP suggests that the Mr 21,000 pre- icamycin-treated ascites tumor cells, the Mr 19,000 pro form cursor contains a signal or pre sequence at the NH2 terminus (lanes 13 and 14) was synthesized but its core-glycosylated Mr as has been shown for other polypeptide hormones (20-23). In 23,000 polypeptide was not. The antibiotic tunicamycin is general, the pre sequence, consisting of 20-30 amino acids, is known to block in vivo formation of the activated carbohydrate- cleaved off cotranslationally by proteolytic enzyme(s), to yield lipid intermediates that are essential prerequisites in the mem- a pro form slightly lower in molecular weight than the brane-dependent glycosylation of glycopolypeptides (25); pro- prepro form. cessing of the prepro form to the pro form, however, is unaf- When hypothalamic mRNA was translated in the presence fected (15). of a microsomal membrane fraction from dog pancreas, the Mr 21,000 precursor was converted to a minor Mr 19,000 and a DISCUSSION prominent Mr 23,000 product identified by anti-Np II (Fig. 2, The data presented here show cell-free ribosome-dependent lane 3) or anti-AVP (Fig. 2, lane 5). That both the Mr 19,000 and synthesis of a polypeptide precursor containing an amino acid 23,000 polypeptides are common precursors to Np II and AVP sequence immunologically identical to that of the nonapeptide was again demonstrated by sequential immunoprecipitation hormone AVP. Sequential immunoprecipitation studies indi- (Fig. 2, lanes 7-10). The recent suggestions (13) that the Mr cate that this oligopeptide exists together with its carrier 18,000 polypeptide might also be processed to a lower molec- Np II as a composite common precursor, supporting recent in ular weight product was most likely due to the less-specific an- vivo studies (7, 18). That the two peptides are part of a common tibodies used in those experiments, which were unable to dis- precursor was also supported by the finding that with both anti- criminate between Np I and Np II. AVP and anti-Np II, electrophoretically identical pro forms, Downloaded by guest on October 2, 2021 Biochemistry: Schmale and Richter Proc. NatL. Acad. Sci. USA 78 (1981) 769 either glycosylated or nonglycosylated, could be obtained. Our 3. Mitnick, M. A. & Reichlin, S. (1972) Endocrinology 91, 1145-1153. data show a Mr 21,000 prepropolypeptide which can be con- 4. Johannson, K. N., Currie, B. L., Folkers, K. & Bowers, C. Y. (1973) Biochem. Biophys. Res. Commun. 53, 502-507. verted into a Mr 19,000 pro form, implying a pre sequence of 5. Bauer, K. & Lipmann, F. (1976) Endocrinology 99, 230-242. approximately 15 amino acids. The latter can also be core gly- 6. Sachs, H. (1969) Adv. Enzymol. 32, 327-372. cosylated to a Mr 23,000 product. With anti-Np II but not with 7. Russell, J. T., Brownstein, M. J. & Gainer, H. (1979) Proc. Natl anti-AVP, another precursor polypeptide (Mr 18,000) could be Acad. Sci. USA 76, 6086-6090. identified that differs from the Mr 21,000 product by approxi- 8. Brownstein, M. J. & Gainer, H. (1977) Proc. NatL Acad. Sci. USA mately 30 amino acids. Incorporation studies with N- 74, 4046-4049. 9. Lauber, M., Camier, M. & Cohen, P. (1979) FEBS Lett. 97, formyl[3S]methionine suggest that the NH2 terminus of the Mr 343-347. 21,000 and the Mr 18,000 polypeptide is intact (13); conse- 10. Acher, R. (1979) Angew. Chem. 91, 905-919. quently, both precursors should be prepro forms. Provided that 11. Guidice, L. C. & Chaiken, J. M. (1979) Proc. Natl, Acad. Sci. USA the two polypeptides are synthesized from one mRNA, the 30 76, 3800-3804. amino acids that lack the Mr 18,000 polypeptide should be lo- 12. Lin, C., Joseph-Bravo, P., Sherman, T., Chan, L. & McKelvy, J. cated at the COOH-terminus of the Mr 21,000 precursor. If this F. (1979) Biochem. Biophys. Res. Commun. 89, 943-950. 13. Schmale, H., Leipold, B. & Richter, D. (1979) FEBS Lett. 108, assumption is correct, the AVP sequence should be at or near 311-316. the COOH-terminus of the Mr 21,000 prepro form. On the 14. Scheele, G., Dobberstein, B. & Blobel, G. (1978) Eur.j. Biochem. other hand, if the Mr 18,000 polypeptide represents another 82, 593-599. discrete Np II-like precursor (18), then the AVP sequence 15. Bielinska, M. & Boime, J. (1979) Proc. Natl Acad. Sci. USA 76, would not necessarily be at the Mr 21,000 COOH-terminus. 1208-1212. The precise arrangements of the two polypeptides Np II and 16. Laude, A. M., Adesnik, M., Sumida, M., Tashiro, Y. & Sabatini, AVP within the precursor will have to await amino acid se- D. D. (1975) J. Cell Biol 65, 513-528. 17. Kessler, S. W. (1975)J. Immunol 115, 1617-1624. quence analysis. 18. Brownstein, M. J., Russell, J. T. & Gainer, H. (1980) Science 207, The data presented here are in good agreement with those 373-378. obtained by direct extraction of neurophysin-like peptides from 19. Schmale, H. & Richter, D. (1980) FEBS Lett. 121, 358-362. neurosecretory granules of the neurohypophysis (9, 18). In 20. Blobel, G. (1978) in Proceedings of the 11th FEBS Meeting, ed. those experiments, larger precursors (presumably pro forms) Clark, B. F. C. (Pergamon, Oxford), Vol. 43, pp. 99-108. having immunoreactive sequences corresponding to neuro- 21. Patzelt, C., Chan, S. J., Duguid, J., Hortin, G., Keim, P., Hein- rikson, R. L. & Steiner, D. F. (1978) in Proceedings of the 11th physins as well as to vasopressin were reported with Mr of FEBS Meeting, ed. Clark, B. F. C. (Pergamon, Oxford), Vol. 47, 20,000-30,000. Recently, a Mr 40,000 potential precursor iso- pp. 69-78. lated under denaturing conditions was described (26). To date 22. Sabatini, D. D., Kreibich, G. (1976) in The Enzymes of Biological we have no evidence to suggest the existence of such a trans- Membranes, Fogarty Conference, National Institutes of Health lation product arising from bovine hypothalamic mRNA. Even (GPO, Washington, DC), pp.171-175. was with calf 23. Davis, B. D. & Tai, P. C. (1980) Nature (London) 283, 433-438. when the translation assay supplemented total 24. Lingappa, V. R., Lingappa, J. R., Prasad, R., Ebner, K. E. & liver tRNA, including termination suppressor species in order Blobel, G. (1978) Proc. Natl. Acad. Sci. USA 75, 2338-2343. to encourage read-through of potentially long mRNA, no high 25. Stuck, D. K. & Lennarz, W. (1977)j. BioL Chem. 252, 1007-1013. molecular weight products were observed (unpublished data). 26. Nicolas, P., Camier, M., Lauber, M., Masse, M. O., Mohring, The existence of composite common precursors has been J. & Cohen, P. (1980) Proc. Natl. Acad. Sci. USA 77, 2587-2591. demonstrated for other neuropolypeptides of which the best 27. Rubinstein, M., Stein, S. & Udenfriend, S. (1978) Proc. NatlAcad. Sci. USA 75, 669-671. studied example is pro-opiocortin (27) or pro-opiomelanocortin 28. Chretien, M., Benjannet, S., Gossard, F., Gianoulakis, C., Crine, (28) containing corticotropin, f3-lipotropin, endorphins, and P., Lis, M. & Seidah, N. G. (1979) Can.j Biochem. 57, 1111-1121. melanotropins (27-32). A common precursor has also been de- 29. Nakanishi, A., Inoue, A., Taii, S. & Numa, S. (1977) FEBS Lett. scribed which includes seven copies of [Met]enkephalin and 84, 105-109. one [Leu]enkephalin (33). These precursors from mammmalian 30. Mains, R. E., Eipper, B. A. & Ling, N. (1977) Proc. Natl, Acad. tissue bear a close resemblance to viral polyproteins in that both Sci. USA 74, 3014-3018. 31. Roberts, J. L. & Herbert, E. (1977) Proc. Natl Acad. Sci. USA 74, exhibit the unusual property of comprising more than one func- 4826-4830. tionally distinct polypeptide (34). In both cases a central role 32. Leipold, B., Schmale, H. & Richter, D. (1980) Biochem. Biophys. is played by the endoplasmic reticulum which appears to define Res. Commun. 94, 1083-1090. a topology for the various processing and modification events. 33. Stern, A. S., Lewis, R. V., Kimura, S., Kilpatrick, D. L., Jones, Although similar in many ways, to date only the viral polypro- B. N., Kojima, K., Stein, S. & Udenfriend, S. (1980) in Biosyn- teins offer evidence of cotranslational proteolysis of the nascent thesis, Modification and Processing of Cellular and Viral Poly- proteins, eds. Koch, G. & Richter, D. (Academic, New York), pp. translation product by viral as well as cellular proteases (35). 99-110. We thank Christiane Schmidt for excellent and skillful technical as- 34. Richter, D., Ivell, R. & Schmale, H. (1980) in Biosynthesis, Mod- sistance and Dr. Richard Ivell for discussions. ification and Processing of Cellular and Viral Polyproteins, eds. helpful Koch, G. & Richter, D. (Academic, New York), pp. 5-13. 1. Lipmann, F. (1973) Acc. Chem. Res. 6, 361-367. 35. Korant, B. D., Langner, J. & Powers, J. C. (1980) in Biosynthesis, 2. Kleinkauf, H. & Gevers, W (1969) Cold Spring Harbor Symp. Modification and Processing of Cellular and Viral Polyproteins, Quant. BioL 34, 805-813. eds. Koch, G. & Richter, D. (Academic, New York), pp. 277-288. Downloaded by guest on October 2, 2021